HR Wallingford Mathematical Model of Groynes on Shingle Beaches A H Brampton BSc PhD D G Goldberg BA Report SR 276 November 1991 Address:Hydraulics Research Ltd, wallingford,oxfordshire oxl0 gBA,United Kingdom. Telephone:0491 35381 Intemarional + 44 49135381 relex: g4gsszHRSwALG. Facstunile:049132233Intemarional + M 49132233 Registeredin EngtandNo. 1622174 This report describes an investigation carried out by HR Wallingford under contract CSA 1437, 'rMathematical- Model of Groynes on Shingle Beaches", funded by the Ministry of Agri-culture, Fisheries and Food. The departmental nominated. officer for this contract was Mr A J Allison. The company's nominated. project officer was Dr S W Huntington. This report is published on behalf of the Ministry of Agriculture, Fisheries and Food, but the opinions e>rpressed are not necessarily those of the Ministry. @ Crown Copyright 1991 Published by permission of the Controller of Her Majesty's Stationery Office Mathematical model of groSmes on shingle beaches A H Brampton BSc PhD D G Goldberg BA Report SR 276 November 1991 ABSTRACT This report describes the development of a mathematical model of a shingle beach with gro5mes. The development of the beach plan shape is calculated given infornation on its initial position and information on wave conditions just offshore. Different groyne profiles and spacings can be specified, so that alternative gro5me systems can be investigated. Ttre model includes a method for dealing with varying water levels as the result of tidal rise and fall. CONTENTS Page 1. INTRODUCTION I 2. SCOPEOF THE UODEL 3 2.t Model resolution and input conditions 3 2.2 Sediment transport mechanisms 6 2.3 Vertical distribution of sediment transport q 2.4 Wave transformation modelling L0 3. DESCRIPTIONOF THE MODEL L0 3.1 Plan shape ehanges 11 3.2 Beach profile changes L4 3.3 Calculation of the distribution of cross-shore transport 15 3.4 Representation of groynes l9 4. INITIAL RESULTSAND FUTURS CALIBRATION 2I 5. REFEPJNCES 25 FIGURES 1 Continuity equation 2 Model of shingle beach profile 3 Distribution of longshore transport 4 Tidal levels at Hythe - probabilistic description 5 Simplified beach with groyne 6a Beach development for 3 gro5me system (0-l year) 6b Beach development for 3 gro5me system (1-10 years) 1. INTRODUCTION Around much of the coastline of England and Wales the natural defences against erosion and flooding are shingle beaches. In the south and east of England i.n particular, from L5rmeBay to The Wash, shingle beaches defend often heavily populated, low-lying areas. The problems of erosion and flooding here are all the greater because the underlying geology is of sedimentary rock and both downward settling of the land mass and eustatic changes in sea levels contribute to a rather rapid submergence of the nearshore seabed, reaching 300run/century in sone areas. Ttre shingle j-tself is typically of chert or flint, with median partiele dianeter between 5nunand 50rrn, and is often a relict of past geological ages. Most of the shingle on the beaches of south-eastern England, for exarnple, was originally deposited by the action of peri-g1acial rivers before the end of the last Ice Age. Although shingle is added to beaches by erosion of cliffs or the nearshore seabed in some areas, the supplies of fresh material are generally small and diminishing as coastal cliffs are defended. A wide, healthy shingle beach is a considerable asset to a coastal conrnunity. Despite the coarseness of the material, such beaches have been popular with tourists since sea-bathing gained royal patronage in the Regency period. Many of the earliest coastal engineeri.ng works on shingle beaches were designed from a recreational viewpoint. More pressing nowadays is the need to prevent inundation of the land behind the beaches, and to prevent erosion or landward mi-gration of the beaches themselves. The traditional method of management of shingle beaches in the IIK has been to install closely-spaced vertically-faced timber groynes, usually of tropical hardwood, to prevent movement of the coarse beach material alongshore. Often, in addition, vertically faced walls of concrete or masonry were built at the crest of the natural beach to prevent occasional overtopping and to support a promenade or roadway along the coast. Erosion problems "downdriftrr of such works were often oqperienced, due to the interruption of the natural transport of beaeh material, and such problems were often tackled using very similar gro5mes and seavalls. In more recent times, alternative or complementary methods of managing shingle beaches have been used. Uechanical re-cycling or bypassing of shingle has been carried out for decades. As an example, the shingle which accretes against the western side of the training waIl at Rye Harbour in Sussex has been placed into trucks and returned to the beach further west (Pett) for many years. Recent innovations have allowed major shingle beach nourishment schemes to be carried out economically, such as at Seaford in 1987 when material was pumped directly ashore through a floating pipeline by the dredger which gathered the material from an offshore area. In some situations the use of active beach management techniques such as re-cycling or periodic re-nourishment will replace the use of groynes. In many more cases, however, gro)rnes will still have a useful role to play in stabilising shingle beaches. In order to deeide on an appropriate management strategy, a numerical model which can simulate beach development including the effects of alternative groJme fields as well as nourishment and recycling is a valuable design aid. It is this reasoning which 1ed to the present research study. 2. SCOPEOF THE MODEL The modelling of beach changes in the vieinity of coastal- structures is beset with difficulties. On one hand, if a model is to be of practical use to a coastal engineer it must be capable of producing accurate predictions of beach changes over a period of years whilst being economical and simple to run. On the other hand, the various interactions between !{aves, tides, coastal structures and sediment transport are often extremely conplex. Even a simplified representation of some of these processes may require very large computational effort coupled with detailed information or measurements from each site to provide a satisfaetory validation or calibration. Because of these eontradictory requirements, it is necessary to make decisions on the scope of the model at an early stage. Some physical processes can be excluded because they are unimportant, whilst others may have to be simplified or ignored because they cannot be model1ed. This chapter focuses on the factors which are important in the development of a gro5med shingle beach, and on the selection of the processes which have been incorporated in the rnodel. The principal objective of the model is to predict beach changes, over a period of years, under different management strategies including different gro5me Iayouts. 2.1 Model resolution and input conditions An important consideration in setting up a numerical model- is the required resolution, and how that resolution will influence the required input conditions. In many areas of the lJK, groSrnes on shingle beaches are built at spacings of 50rn or less. As a result, it is necessary to set a fine alongshore spacing (5-1Om) in order to represent adequately the beach plan shape within each bay. Another conseguence of such narrow compartments is that the orientation of the beach contours can change very rapidly during a storm. To ensure that the calculated breaking wave conditions and the beach morphology do not rtclashr and cause numerical instabilities, it becomes necessary to use small timesteps in the model (about 1-3 hours or less in violent storms). When the requirement is to study the evolution of a many-groyned beach over several years, the number of timesteps and positions along the shore both become 1arge. This influences the conplexity of the modelling that can be carried out whilst stil1 retaining an economical design method (see sections 2.2 - 2.4 below). It is also important to bear in rnind from the outset that any numerical model will depend on the accuracy of its input data in producing accurate results. Careful consideration has to be given therefore, to what information the coastal manager can be expected to provide, especially when he or she is interested in predi.cti-ng of future beach development as well as explaining past beach changes. Specification of an initial beach plan shape, the position of seawalls and the positions and shapes of existing or proposed gro5mes is usually straightforward. Simi-Iarly, information on beach material and beach profiles can usually be obtained quickly or, in the case of a proposed new beach, estimated reasonably aecurately. More di-fficult to specify are the time varying input conditions which produce changes in the beach. The main time-varying input information required for the BEACHPLANmodel is wave data. Four different methods of prescri.bing wave conditions are possible, as follows: 1. A single rmorphologicalr average wave condition assumed to persist throughout the whole period for which the model is run. 2. Several ttypicalr wave conditions, arranged in a pre-specified sequence which repeats over the period of the model run. 3. A sequence of wave conditions chosen from a directional wave clirnate specified by the probability of any combination of wave height, period and direction occurring. lhe sequence has to be chosen by some random-choice algorithm. 4. A sequence of wave conditions (wave heights, periods and directions) which has been recorded or calculated retrospectively (ie hindcast) from recorded wind data. In many situatj-ons it is found that initial runs of the model are best carried out with a single wave condition (option 1) followed by sophisticated wave input conditions at a late stage.